Cyclin Dependent Kinase-2
Created by Ryan Hunter
General function
Human cyclin dependent kinase 2 (CDK-2) is a Ser/Thr kinase that functions in cell cycle progression. As with other members of the CDK family, it forms heterodimers with cyclin (specifically cyclin A and E for CDK-2) and works to initiate the next step in the cell cycle. CDK-2 is activated only during S and G2 phase, with its most important function in activating DNA replication. In terms of cell cycle activation, CDK-2 seems to have fully redundant function compared to CDK-1; that is to say, if CDK-2 is missing in mammalian cells, they are still able to proceed through the cell cycle. Furthermore CDK-2 is involved in a regulatory pathway involving Cdc6 and ATRIP which monitors CDK-1 activity. Recent research has shown that it also has crucial-and in this case non-redundant-function in cell cycle arrest in response to DNA damage at the G-2 and S phase junctions (Chung, & Bunz, 2010). Overall, CDKs are regulated by transcriptional and post-translational mechanisms which orient them to provide correct timing/order in the processes of cell growth and division. The activated CDK-2 will be so when it is phosphorylated (Thr-160 residue) and is bound by a regulatory cyclin subunit (Brown, Noble, et. al, 1999). This activated subunit can then be inactivated by phosphorylation at two residues, T-14, Y-15. Finally, some research has shown that an additional 20K protein can also serve as a negative regulator of CDK-2 (Gu et al, 1993). These regulatory mechanisms will be discussed further below.
Secondary Structure
CDK-2 is an alpha and beta (a + b) protein, meaning that it is composed of several antiparallel beta sheets and alpha helices which are segreagated; CDK-2 has a characteristic kinase fold with an N-terminal domain made of mainly beta-pleated sheets and a carboxyl domain consisting of primarily alpha-helices. Overall the secondary structure is 34% helical (13 helices, 102 residues), and 15% beta strands (11 strands, 47 residues). This leaves 149 residues that exist as turns, random coils, and other non-specific structural segments (Brown et al, 1999). In essence, the protein is composed of two domains;
domain 1 exists between residues 1-84. It is an alpha beta domain, also described as a two-layer sandwich, and is active as a phorphorylase kinase. This region is dominated by beta strands but an alpha helix is present. The beta sheets are concentrated in this domain, between residues 4 and 81 (note, there are two other beta strands at residue 129 and 141). There are two important activating residues here (Thr-14 and Tyr-15) that will be discussed in further detail below (Wissing, J et al, 2007).
Domain 2 exists between the remaining residues 85-298 and is composed of mainly alpha helices which are described as an orthogonal bundle (referring to the relative angle displacement between the constituents). These alpha helices are most prevalent through the rest of the protein between residues 87 and 282 (again note, an extra-regional alpha helix also exists at residue 46). A few short beta strands are located in this domain as well. Another activating residue, Thr-160 is located in this domain and will be discussed below. As evidenced from the image, the two domains are more or less clearly segregated, with limited overlap in structural similarities between the two regions (Brown et al, 1999).
Functionally Important Residues
Asp-127 is an important site as a proton acceptor, and spatially it exsists in the vacinity of Lys-33, Asp-86, and Asp 145 which compose an
ATP binding site. Residues 10-18, 81-83, and 129-132 should also be noted as assisting in creating a pocket where ATP can bind, also called the
nucleotide binding region.
Thr-14 and Tyr-15 are adjacent, important phosphorylable residues which compose a portion of the ATP-binding domain just discussed; when they are phosphorylated, the enzyme is inactivated. Although Thr-160 is generally noted as the crucial residue in terms of activating this protein, these two grant an additional level of post-translational modification by offering a way to inactivate the protein even when T-160 is phosphorylated. This property has been expressed experimentally: when the Thr-14 or Tyr-15 is changed to an alanine residue, there is a two-fold increase in enzyme activity (Gu et al, 1992).
Thr-160 is the most significant residue in terms of activating CDK-2. It is similar to the previous two in that it gets phosphorylated, but it is functionally reversed; when Thr-160 is phosphorylated, the enzyme activity increases. In relation to the experiment mentioned, when Thr-160 was changed to A-160, enzyme activity was abolished (Gu et al, 1992). It is interesting to note that while T-160 is in close proximity to T-14 (12.2 Angstroms) and Y-15 (8.2 Angstroms), it is the residue furthest exposed and therefore most accessible to phosphorylation. Residues 145-155 are known as the "activating segment" of CDK-2 because of their role in phosphorylation and subsequent activation of the protein. Phosphorylation of Thr-160 occurs more readily when CDK-2 is bound by cyclin A (Jeffery et al, 1995).
The
C-helix containing the PSTAIRE motif (PSTAIRE are the one letter abbreviations for amino acids 45-51) is an alpha helix located in the middle of domain 1 of CDK-2. This helix has important stabilizing interactions with K-33 (specifically interacting with E-51 of the PSTAIRE motif), which is important for directing the positioning of ATP. Notice the relative positioning of the PSTAIRE helix and the K-33 molecule relative to the
ATP pocket. The geometric layout of the PSTAIRE motif and the proximity of the K-33 residue alone suggest a role in ATP direction. Once the protein is activated, T-160 is then buried in the protein away from its position exposed to the solvent (Brown, Noble, et al, 1999). Specifically, the PSTAIRE region interacts with the cyclin-A box when the ligand is bound to provide for heterodimer activation (Jeffrey et al, 1995).
Important CDK-2 Interactions
Catalytic Site: As noted, D-127 is a proton acceptor and is also the crucial residue in the catalytic site for CDK-2. It is here that binding of cyclin A-E can take place. When cyclin-A binds CDK-2, it remains structurally the same, but CDK-2 changes conformation to create an ATP recognition site, as shown in the discussion/slides above. Other local residues composing the catalytic site include Q-131, K-129, N-132, and T-165 (Brown, Noble, et al, 1999).
The
cyclinA-CDK2 interface more broadly is formed by interlocking various segments from each protein. The significant regions on CDK-2 inclue the PSTAIRE helix, T-loop (res 146-166), portions of the N-terminal beta sheet, and the C-terminal lobe. The PSTAIRE helix has already been discussed. The T-loop generally blocks the catalytic cleft from outside molecules interacting when CDK-2 is not bound by cyclin A and in the inactive conformation. From cyclin A the significant motifs are helices a3, a4, a5, and the N-terminal helix. Overall, the interaction is centered around PSTAIRE interaction with the cyclin A box. CDK-2 is activated by a PSTAIRE conformational change which realigns the active site residues, and causes a motion in the T-loop which removes the blockage of the catalytic cleft. Finally, this conformational change exposes Thr160 which makes it more easily phosphorylated (Jeffery et al, 1995).
Magnesium Ion Binding: D-127 is also noted as an important residue becuase this is where the Magnesium ion cofactor binds. Again it is crucial to notice the close proximity of the
catalytic site to the ATP binding region; this also makes logical sense indicating the necessity of a phosphate group transfer to fuel catalysis. The cofactor is also locally present to assist in electron/molecular exchange. *For the two slides of this section, N-terminal residues 1-44 have been removed for easier viewing*.
ATP Binding: As mentioned above, CDK-2 necessarily binds ATP. Amino acids K-33, D-86, and D-145 are the crucial components of CDK-2 nucleotide binding. The ATP-binding site is situated at the domain interface (Brown, Noble, et al, 1999).
CDK-2 Compared to other proteins
An obvious place to look for sequence and tertiary homogeneity is within the family of cyclin dependent kinase proteins.
Cyclin dependent kinase 5 (PDB ID: 1H4L) is a protein structured much like CDK-2 that carries out a similar function. CDK-5 is known for its important function in regulating the cell cycle in neural cells specifically. When its function is inhibited or decreased neural degeneration will often result. CDK-2 is composed of 299 amino acids, similar to CDK-5 which has 292. CDK-2 has a molecular weight of 34002.8 Daltons, compared to that of CDK-5, 33364.8 Daltons. As for secondary structure, CDK-2 is composed of 34% alpha-helices and 15% beta-sheets; CDK-5 is 35% and 13% respectively (Brown, Noble, et al, 1999) and (Tarricone et al, 2001) . CDK-2 has an isoelectric point of pH 8.8 and CDK-5 has a comparable value of 7.57 ("Expasy."). In terms of tertiary structure, the spatial layout of CDK-2 and CDK-5 is quite similar, with clusters of alpha helices and beta sheets concentrated in separate domains.
Cyclin dependent kinase 7 (PDB ID: 1UA2) is most similar to CDK-2 by this measure of similarity in protein folding/subunit spatial arrangement (Lolli et al, 2007).